Dead Ends or Dividend: A Revenue-Weighted Triage for Retired UK Product URLs
UK retailers retire products constantly: seasonal ranges end, suppliers change, sizes sell through and long-tail lines are quietly discontinued. The operational instinct is to redirect every dead product URL to the nearest category and move on. That is tidy for the redirect file and often poor for search. A category page rarely answers the specific query that brought a shopper to a product URL, and a blanket redirect can hide useful demand, break conversion paths and weaken the evidence available to search and AI systems.
In 2026, the better approach is a revenue-first redirect map: a repeatable way to score retired URLs before choosing between a permanent redirect, a gone response and a refreshed landing page. The aim is not to preserve every URL. It is to allocate redirect effort where residual demand, historical revenue, replacement fit and citation risk justify it.
Why blanket category redirects fail
A category redirect is a compromise. It tells a search engine that the old product page has moved, but it does not tell the shopper which replacement is closest to the retired item. For a discontinued running shoe, a redirect to the entire footwear category may satisfy a crawler and frustrate a buyer. For a spare part, it may be worse: the category page does not confirm compatibility, so the user returns to search.
There is also a measurement problem. Once every dead URL lands on the same category, analytics can no longer distinguish between genuine category demand and traffic that arrived through a mismatched product redirect. Conversion rate falls, but the cause is buried. AI-assisted search adds another layer: answer engines and shopping agents may still cite or surface an old product URL, and a category page that does not address the original product can create a poor citation outcome.
A four-factor score for retired URLs
A practical model uses four factors, each scored consistently and reviewed by the same owner. The weights below are a starting point, not a universal law; retailers with strong margin data or heavy paid support may adjust them.
- Search demand (30%): impressions, clicks and query themes from Search Console over the previous 90 days, plus internal site search where available. Score 0 to 5.
- Historical revenue (30%): attributed revenue and orders over the previous 12 months, preferably margin-adjusted. Score 0 to 5 rather than using raw revenue, which can let a few high-ticket returns distort the map.
- Replacement fit (25%): how closely an in-stock successor matches the retired product on use case, specification, price band and compatibility. Score 0 to 5.
- AI-citation risk (15%): evidence that the URL is referenced by external pages, answer engines, shopping feeds or high-authority mentions. Score 0 to 5. This is not about chasing citations; it is about avoiding a clearly wrong answer when a machine-driven surface points at a dead page.
The weighted score informs the action, but it should not automate it. A high demand score with poor replacement fit is a signal to refresh a landing page or build an alternatives page, not to force a 301. A low score across all four factors is a candidate for a 410.
The decision matrix
Use the score alongside two hard checks: replacement availability and intent match. A sensible matrix looks like this:
| Condition | Preferred action |
|---|---|
| High weighted score and replacement fit of 4 or 5 | 301 to the closest replacement product |
| High demand but replacement fit below 4 | Refresh as a discontinued-product landing page with alternatives |
| Low demand, no revenue and no credible replacement | 410 Gone |
| Seasonal product with recurring demand | Keep the URL live, update availability and add a waitlist or successor link |
| Long-tail spare part with support value | 301 to the compatible part or a compatibility hub |
A 410 is not a failure. It is a clean signal when a product has no successor and no meaningful residual demand. The mistake is using 410 as a default for operational convenience while ignoring URLs that still earn links, impressions or assistant citations.
A worksheet that survives handover
The redirect map should be an auditable asset, not a one-off export. A worksheet with the following columns is sufficient for most UK catalogues:
URL | status | 90-day impressions | 90-day clicks | 12-month revenue | query theme | replacement SKU | replacement fit (0-5) | citation risk (0-5) | weighted score | action | owner | review date
Add a second tab for redirect rules and a third for evidence links. The evidence tab matters: a screenshot of a Search Console query cluster or a reference to a supplier discontinuation notice is more useful to a successor than a bare score.
For example, a discontinued waterproof jacket with steady branded queries and a close successor should receive a 301 to that successor, not to the outerwear category. A retired colourway with no specification difference may redirect to the same product in another colour, provided the page makes the change clear. A low-demand novelty item with no replacement and no citations can be retired with a 410.
Seasonal, long-tail and discontinued ranges
Seasonal ranges need special handling. Redirecting garden furniture to a generic outdoor living category every October trains users and crawlers to expect instability. Where demand recurs, keep the URL live with clear availability messaging and a link to the current season or successor range. If the range is genuinely retired, redirect to the closest replacement only when the fit is credible.
Long-tail spare parts are often undervalued by revenue-only models. A low-revenue component may carry high support value and strong compatibility intent. Redirect it to the compatible part or a compatibility hub rather than a broad accessories category. Discontinued ranges with a loyal following may justify a refreshed landing page that explains the discontinuation and lists alternatives; this preserves the query path without pretending the original product is available.
Governance before automation
Redirect maps decay. Products return, replacements sell out and new URLs are created without updating the rules. Assign one owner, review the map quarterly, and monitor 404 and soft-404 reports after every catalogue change. Avoid redirect chains, keep the final destination canonical, and test a sample of high-score URLs in a staging environment before deployment.
The revenue-first discipline is simple: do not let a dead URL receive the same treatment as every other dead URL. Score the residual value, choose the action deliberately, and document why. That is how a redirect map becomes a commercial asset rather than a housekeeping exercise.
